[kszx]

It may be true that this particular sample is too small; I haven't looked at the experiment in detail. However, just pointing out that n<=20 per group isn't enough of an argument to justify this statement. Significance is about more than just the sample size. The idea that it's impossible to get significant statistical result with small samples is a common misconception.

[thanatosmin]

True. The study provides the requisite observation variances and effect sizes for this analysis as well.